00:00
Hi my name is Daryl Hayward from the portfolio marketing team at Pure. And today we're gonna talk a little bit about the effective used capacity also known as E. U. C. So effective use capacity is the daily average of host written data on pure storage arrays. And it's used to calculate the billing for
00:28
evergreen one evergreen flex as well as cloud block store. So Pure does not simply bill on provision capacity for these services. Even though this may seem like a simpler approach. It may not be fair to you since all volumes are thin provision instead we build for unique host written data
00:54
which lets you take advantage of unique storage features like snapshots and volume clones without incurring huge costs. So what I'm gonna do now is walk through a few examples of the U. C. To show you how it works. Alright so first we're gonna take a look at an independent unprotected volume where there's no snapshots or replication to consider. So when the host writes data to the volume,
01:22
this data is represented as chunks of continuous data called an extent. Alrighty. So so let's let's take an example of where your you see is equal to three member bites. I'll refer to those as M. I. B. From now on. Okay this is the amount of data where the host specifically writes data to the volume
01:47
and is measured before any data reduction occurs. So now let's say the host overwrites one of the extents since we are still only storing three M. I. B. U. C. Remains at three because we're still storing the same amount of data in the volume. But what if the host unmapped and deletes one of the extents?
02:15
Well now Ari you see has decreased to to M. I. B. So now what if we cloned this volume? Right so I'm gonna go back to the original volume with three extents and I'm gonna clone this volume are you see actually goes back to three you know to match the original volume and it remains at three. And that's because you know we only make
02:47
changes to Eu see if there are changes in the original volume. Let's say we overwrite one of the extents in the E. U. C. And then we add another as well. So in this case the E. U. C increases Yeah 25 M. I B.
03:10
And that's because we still have to store the original three extents and then also account for the two changes as well. And the same would apply to any changes that we would make in the clone volume as well. Alright so so we just take a look at at clones but now let's take a look at snapshots which are actually very similar. Um We're gonna go back to our original
03:38
volume here with our E. U. C. At three a.m. I. B. Okay so now if we take a snapshot of this volume here. Sorry you see is not affected. We only took a virtual snapshot of of the volume here. So no changes have been written to the actual
04:05
volume but so once we start writing new data to the volume let's say we write a new extent here Then are you see increases because we're storing additional data on the original volume in this extent increases from three 2 4 More data stored on the original volume You see goes from 3-4. So now let's write some more data and take another snapshot.
04:33
So we're gonna write two more extents and then take a snapshot of this. So now our EU. C increases to six. So what if we were to delete one of the extents from the original volume? So actually our E U. C wouldn't change. Right because we still have to account for the six unique extents in the snapshot.
05:07
So what if we were to overwrite one of the extents in the original volume? So are you see would then actually increase 27 because you see is all about how much data is written to the array. So the number of snapshots doesn't necessarily impact the effective use capacity is just the amount of data that has to be stored. So I'm gonna draw a couple more snapshots in
05:34
here. Right boom boom boom. And then just to keep it consistent will represent the overwrite the same way and we'll do it again here. Another snapshot and then again keeping it consistent will do the overwrite the same way.
06:01
So now um with that guess what the U C. Doesn't change because all I did was make snapshots of the of the volume of the original volume here, so we're still at the U C. Of seven. Alright, so originally we talked about clones, then we talked about snapshots and now we're gonna talk about replication and we're gonna start with asynchronous replication.
06:23
So a synchronous replication, so when we first start replicating, you know, all the data is copied to a secondary array. So since we have two physical copies of the data, you see then
07:02
is six so to physical copies of the data, E U C is equal to six M I B let's write two additional extents to the volume. And so are you see increases by a factor of one? So we added two extents. Sorry, you see, now increases 28. So it's not until changes have been snapped and replicated to the secondary array is when you're e u c is affected.
07:38
So now in this case sorry, you see increases 2 10, Of course the more raise the more effective use capacity so you can see how your r p o might impact your you see, and that is because you see is calculated on an average daily basis. Alright, so the last example, what's it's kind of like example, three be really straightforward.
08:05
This is actually active cluster. So I'm gonna draw couple of volumes here. So the reason I say this is pretty simple is because with active cluster, both of your volumes will be identical. So the E U C is always double. So again, in this case the U C would be five plus five is 10.
08:41
Okay, so those were a couple of quick examples of how you see works. And so next I'm gonna show you are pure one management tool which is our Ai driven data services platform for storage management and it's provided to all of our customers at no additional cost. And you see can easily be tracked by looking at the subscriptions page or at the analytics section under capacity for more
09:06
information and documentation on the you see feel free to go to our pure storage dot com website but also contact your account team, you know, your your stories account team has a lot of information um and they'll be able to help you understand the you see a lot better and they will also be able to help you understand how to uncomplicated your data storage forever.
